WO2006048988A1 - Système de recherche de données de contenu - Google Patents

Système de recherche de données de contenu Download PDF

Info

Publication number
WO2006048988A1
WO2006048988A1 PCT/JP2005/018353 JP2005018353W WO2006048988A1 WO 2006048988 A1 WO2006048988 A1 WO 2006048988A1 JP 2005018353 W JP2005018353 W JP 2005018353W WO 2006048988 A1 WO2006048988 A1 WO 2006048988A1
Authority
WO
WIPO (PCT)
Prior art keywords
search
range
attribute value
impression
music
Prior art date
Application number
PCT/JP2005/018353
Other languages
English (en)
Japanese (ja)
Inventor
Tadashi Yoshida
Hitoshi Araki
Yoshiteru Kawasaki
Masato Yuda
Kazuya Nomura
Original Assignee
Matsushita Electric Industrial Co., Ltd.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Matsushita Electric Industrial Co., Ltd. filed Critical Matsushita Electric Industrial Co., Ltd.
Priority to US11/665,663 priority Critical patent/US7743049B2/en
Priority to JP2006542314A priority patent/JP4659755B2/ja
Priority to EP05790249.6A priority patent/EP1832987B8/fr
Priority to CN2005800368296A priority patent/CN101048776B/zh
Publication of WO2006048988A1 publication Critical patent/WO2006048988A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2452—Query translation
    • G06F16/24522—Translation of natural language queries to structured queries
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/40—Information retrieval; Database structures therefor; File system structures therefor of multimedia data, e.g. slideshows comprising image and additional audio data
    • G06F16/48—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually

Definitions

  • the present invention relates to a content data search apparatus for searching content data.
  • an attribute value indicating an attribute is assigned to each content item. Also, the search term is converted into information indicating the range of attribute values corresponding to the search term. Then, the content data corresponding to the attribute value included in the range of the attribute value is output as the search result.
  • Impression words are qualitative words and phrases that express an impression on content such as music such as "loose style”.
  • an impression value which is an index value representing an impression of the music
  • the impression word input as a search word is converted into the range of the corresponding impression value.
  • the music data corresponding to the impression value included in the range of the impression value is music data corresponding to the search.
  • the music search system is applied to, for example, music reproduction software for car navigation devices and personal computers.
  • Patent Document 1 Japanese Patent Application Laid-Open No. 2004-86189
  • the search target is a music group narrowed down by a specific artist name
  • many of the music become similar impressions.
  • the impression values of the songs become close to each other in the absolute position in the space where the impression values exist. Therefore, when a search is performed using an impression word, many songs are selected for a particular impression word, but there is no corresponding song for other impression words. Become.
  • impression word music retrieval systems do not take this into consideration.
  • the present invention has been made in view of the above problems, and is a content data search apparatus for searching content data associated with an attribute value indicating a content attribute using a search term, which is a content to be searched
  • An object of the present invention is to provide a content data search device capable of appropriately narrowing down search targets even when distribution of group attribute values is biased.
  • a content data search device of the present invention is a content data search device for searching content data associated with an attribute value indicating an attribute of content by a search term,
  • a conversion table in which attribute value storage means for storing values, search terms, and search term ranges that are ranges of attribute values corresponding to the search terms are associated.
  • the attribute value storage means power or the conversion table power is changed according to the distribution of the attribute value corresponding to one content data to be searched, and the search term range read out.
  • the changing means for changing the relative relationship between the distribution range of the attribute value and the search word range, the receiving means for receiving the input of the search word, and the changing means
  • search means for searching for content data corresponding to the search term by searching for an attribute value included in a search term range corresponding to the search term accepted by the search term.
  • the content data search device of the present invention changes the relative relationship between the distribution range of attribute values and the search word range according to the distribution of attribute values of content data to be searched. Can. Furthermore, in the relative relationship between the distribution range of the attribute value and the search word range after the change, the content data can be searched by the input search word. Therefore, even if the distribution of the attribute values of the content data to be searched is biased, the search target can be appropriately narrowed.
  • the changing unit changes the search term range read out from the conversion table power by reducing or moving so as to be included in the distribution range
  • the search The means searches for content data corresponding to the search term by searching an attribute value included in the search term range changed by the change means, corresponding to the search term accepted by the acceptance means.
  • the changing means changes the attribute value read out from the attribute value storage means so as to expand the distribution range, and the search means corresponds to the search word accepted by the accepting means. Search the content data corresponding to the search term by searching the attribute value changed by the changing unit, which is included in the search term range Chiyoi as a be that.
  • the relative relationship between the distribution range of the attribute value and the search word range can be changed. That is, it is possible to appropriately narrow down the search target.
  • the present invention is realized as a method in which the characteristic component of the content data search device of the present invention is a step, or is realized as a program including those steps, or the program is stored.
  • the user when the content data is music data and an impression value is associated as an attribute value, the user is designated as a search term according to the distribution of impression values of the music to be searched.
  • the relative relationship between the range of impression words and the distribution range of impression values can be changed.
  • the search target is a music data group narrowed down under a specific condition and the distribution of impression values is biased, it is possible to perform a search that reflects the intention of the user who performs the music search. , You can get the appropriate number of songs.
  • FIG. 1 is a block diagram showing a functional configuration of a music search device according to an embodiment of the present invention.
  • FIG. 2 is a view showing an example of impression value data stored in an impression value database.
  • FIG. 3 is a conceptual view showing an example of an impression space in the embodiment of the present invention.
  • FIG. 4 is a diagram showing an example of the data configuration of the impression word conversion table stored in the conversion table storage unit.
  • FIG. 5 is a conceptual diagram showing an example of changing the range of the impression value by the range changing unit.
  • FIG. 6 is a flowchart showing an example of the process flow of the music search device of the embodiment of the present invention.
  • FIG. 7 is a conceptual view of an impression space showing the difference between a search result by the conventional music search system and a search result by the music search device of the present embodiment.
  • FIG. 8 is a functional block diagram showing a functional configuration of the music search device provided with the impression value changing unit.
  • FIG. 9 is a diagram showing an example of the process flow of the music search device provided with an impression value changing unit
  • FIG. 10 is a search result when the distribution range of impression values is not changed;
  • Fig. 11 is a conceptual diagram showing an example where only the radius of the impression word range is changed in the impression space.
  • Figure 12 shows the search results for a certain search word range and the search results for a search word range that has been changed according to the distribution of age values, when music data is searched with fuzzy words related to the age. Diagram of the age space showing the difference between
  • FIG. 1 is a block diagram showing a functional configuration of the music search device 100 according to the embodiment of this invention.
  • the music search device 100 shown in FIG. 1 is an example of the content data search device of the present invention, and is a device for searching music data by an impression word designated as a user search word.
  • the music search device 100 has a music search function such as an audio device capable of storing and reproducing a large amount of music data, a home server, an in-vehicle AV server, a portable music player, and a mobile terminal. It is realizable as an apparatus which it has. It is also possible to realize the functional configuration of the music search device 100 as software and incorporate it into a music reproduction application program for a personal computer.
  • the music search device 100 includes a search result storage unit 101, a music data storage unit 105, a reception unit 107, a conversion table storage unit 108, a range change unit 109, and a conversion unit.
  • a search unit 111, an impression value calculation unit 106, and an output unit 120 are provided.
  • the music data storage unit 105 includes a music database 102, a music information database 103, and an impression value database 104! /.
  • the search result storage unit 101 is a storage device for storing information on the result of the search executed in the music search device 100.
  • the information related to the search result is information such as a music ID of music data corresponding to the search and an impression word which is a search term used for the search.
  • the song ID is an identifier for identifying a song.
  • the music data storage unit 105 is a configuration unit that stores a plurality of music data, music information, and an impression value in association with each other. Specifically, the music data, the music information and the impression value are stored in the music database 102, the music information database 103 and the impression value database 104, respectively. Information and data stored in each database are linked by music ID.
  • the music data stored in the music database 102 is a music signal of music.
  • a plurality of music signals are stored in the music database 102. These music signals are associated with the music ID, and by designating the music ID, data of a desired music signal is called out. Also, these music signals are, for example, waveform data by linear PCM, and compressed data such as MP3 and AAC.
  • the music information stored in the music information database 103 is information on music such as artist name, album name, music name, genre, release year, composer name and the like.
  • the music information of each music is stored in association with the music ID.
  • the impression values stored in the impression value database 104 are calculated by the impression value calculation unit 106. Impression values are stored in association with music IDs.
  • the impression values stored in the impression value database 104 will be described later with reference to FIG.
  • the impression value is an example of the attribute value in the content data search device of the present invention
  • the impression value database 104 is an example of the attribute value storage means in the content data search device of the present invention.
  • the reception unit 107 is a processing unit that receives an impression word specified by the user.
  • impression word in the present embodiment is a term that expresses an impression of music, and is a term that expresses music intuitively, such as “Norinori”, “Pop”, “Healing”, and the like.
  • the conversion table storage unit 108 is an impression word conversion table in which an impression word is associated with information indicating a range of impression values corresponding to the impression word (hereinafter, “impression word range” t ⁇ ⁇ .) It is a storage device to store.
  • impression word range a range of impression values corresponding to the impression word
  • Range changing unit 109 is a processing unit for solving the problems of the conventional music retrieval system based on impression words, acquires impression word conversion table from conversion table storage unit 108, and searches for music to be searched. It is a processing unit that changes the content of the acquired impression word conversion table according to the distribution of the impression value.
  • the conversion unit 110 uses the impression word conversion table changed by the range change unit 109, and a range of impression values corresponding to the impression word designated by the user (hereinafter referred to as “impression word range”). Show It is a processing unit that converts it into The impression word range is an example of a search word range in the content data search device of the present invention.
  • the search unit 111 is a processing unit that searches for music data corresponding to an impression word designated by the user by searching for an impression value included in the impression word range.
  • the output unit 120 is a processing unit for outputting the result of the search performed by the search unit 111 to the user.
  • the search result output from the output unit 120 is presented to the user by a display device or the like (not shown).
  • the impression value calculation unit 106 is a processing unit that calculates an impression value of music from the music data stored in the music database 102. Specifically, the physical feature quantities of the music signal input from the music database 102 are analyzed, and an impression value obtained by weighting and adding the feature quantities is calculated.
  • the physical features include: spectral edge excess (degree of spectral change between frames of a predetermined length of time), average number of pronunciation (pronunciation frequency of sounds pronounced in music), pronunciation non-generation Periodicity (degree of non-periodicity of the sound produced in the music), beat cycle (time length equivalent to quarter note of music), beat cycle ratio (period equivalent to quarter note of music) Sound intensity), beat intensity ratio (ratio of first beat intensity to second beat intensity), etc. are analyzed.
  • the impression value calculation unit 106 performs predetermined weighted addition on these feature quantities obtained by analysis to calculate an impression value. Impression values are calculated for one or more types of music. In addition, predetermined maximum and minimum values that can be taken as an impression value are set.
  • the music can be represented as a point in a two-dimensional space.
  • the music is represented as a point in the 5 dimensional space. That is, the music piece stored in the music data base 102 has coordinate values reflecting the respective impression values in this space. Therefore, each of the music pieces stored in the music piece database 102 can be distinguished and distinguished by the combination of these impression values.
  • the impression value is an index value that represents an impression of a person, such as feeling, emotion, etc., when he or she listens to music, such as intenseness, liveliness, exhilaration, simplicity, softness, etc. is there.
  • the impression value can be expressed as a linear or non-linear function of each of the plurality of feature quantities.
  • Five types of impression values intense, dynamic, exhilarating, rustic, and soft, music is represented as points in a five-dimensional space.
  • the two are combined into the active factor, the intense and the dynamic, and the three are combined as the emotional factor, the exhilaration, the simplicity, and the softness.
  • the music is represented as a point in a two-dimensional space when the five types of impression values are put together into two impression values of an active factor and an emotional factor.
  • Such a space coordinate system in which the impression values of the music are distributed is hereinafter referred to as an impression space.
  • Impression value is a quantity that indicates the impression of listening to the music. Therefore, when the impression values of a plurality of music pieces are arranged on the impression space, the impression values are laid out according to the similarity of the impression to the music pieces. That is, music with similar impressions is placed close to the impression space.
  • Impression values used in the present embodiment are two types, impression value X, which is an impression value based on the above-mentioned emotion factor, and impression value Y, which is an impression value based on the above-mentioned active factor.
  • FIG. 2 is a view showing an example of impression value data stored in the impression value database 104. As shown in FIG.
  • two impression values are associated with each music ID, and a set of two impression values can be treated as coordinate values of one point in a two-dimensional impression space. . That is, the coordinate value of a certain music can be treated as an index that represents the impression of the music, for example, "There is a feeling of excitement and exhilaration.” Therefore, the coordinate value consisting of one or more impression values is also the "impression value" of the music.
  • FIG. 3 is a conceptual diagram showing an example of an impression space in the present embodiment.
  • the impression space in the present embodiment is two-dimensional as described above, and it is “Soft” as it goes to the right and “Hard” as it goes to the left. It also shows that going up is “Active” and going down is "Quiet”.
  • the two-dimensional impression space has a predetermined size.
  • both of the horizontal axis direction (X axis direction) and the vertical axis direction (Y axis direction) are from -4 to +4. That is, the minimum value is 4 for both the impression value X and the impression value Y, and the maximum value is +4.
  • each of the points in the impression space shown in FIG. 3 corresponds to one music unless there is a music having exactly the same impression value.
  • the points in the impression space shown in FIG. 3 are It can be seen that it is biased toward the center. In other words, it can be said that the music corresponding to the point in this impression space is close and the music of the impression is many.
  • the range changing unit 109 acquires the contents of the impression word conversion table acquired from the conversion table storage unit 108 according to the distribution of the impression values in the impression space. Change
  • FIG. 4 is a view showing an example of the data configuration of the impression word conversion table stored in the conversion table storage unit 108.
  • the impression word conversion table is a table for converting an impression word designated by the user as a search word into information indicating an impression word range.
  • imprint words such as “Kukiki” and “Pop” are set.
  • the user can specify an impression word to be used as a search word from among the impression words set in the impression word conversion table, using an input device such as a mouse (not shown in FIG. 1).
  • the impression word range is as shown in FIG. 4 (X, Y)
  • the impression word range is a range that is within the impression space and that satisfies each equation set in the impression word conversion table.
  • Range changing section 109 in the present embodiment acquires the impression word conversion table shown in FIG. 4 from conversion table storage section 108, and changes the contents thereof according to the method described below.
  • the range of the impression value of the music to be searched is (one 0.9 ⁇ X ⁇ 1.3), one (one 1.6 ⁇
  • the range (circle 1) inside a circle of radius 1 centered on (0.4, -0.2), which is the impression word range corresponding to the impression word "pop system", is this mapping function. It is converted to the range inside the ellipse shown in (Equation 4) using
  • the range changing unit 109 changes the range of each impression word read out from the conversion table storage unit 108 according to the distribution of the impression value. Furthermore, the impression word conversion table to which this change has been made is passed to the conversion unit 110.
  • conversion unit 110 converts the user-specified impression word received by reception unit 107 into information indicating an impression word range.
  • range change unit 109 described above
  • the impression word conversion table changed by the method of is referred to.
  • Information indicating the impression word range obtained by this conversion is sent to the search unit 111. That is, in the present embodiment
  • the modified equation exemplified in the above (Expression 4) is sent to the search unit 111.
  • the search unit 111 searches for music data corresponding to the impression word range changed by the range changing unit 109 which is not in the impression word range set in advance in the impression word conversion table.
  • FIG. 5 is a conceptual diagram showing an example of changing the range of the impression value by the range changing unit 109. As shown in FIG.
  • FIG. 5 shows the impression word range before the change of “pop system” in the impression space
  • the right figure shows the impression word range after the change of “pop system” in the impression space.
  • the majority of the search target music corresponds to the “pop type”, and the narrowing effect by the search is provided to the user You can not do it.
  • the impression word range is changed by the range changing unit 109, the proportion of songs that fall under "pop" is reduced to an appropriate number.
  • the search of the music data by the impression word specified by the user is executed, and the music ID, the music information and the music data of the music corresponding to the impression word are output as the search result.
  • information on search results such as music IDs and impression words used as search terms is stored in the search result storage unit 101 as described above.
  • the search result storage unit 101 is used to execute the next search.
  • FIG. 6 is a flow chart showing an example of the process flow of the music search device 100 according to the embodiment of this invention.
  • the process performed by the music search device 100 includes a process (S201 to S203) related to the creation of the impression value database 104 necessary for the search of music data and a process (S204 to S210) related to the execution of the music data search. It is divided into two.
  • music data is stored in the music database 102 by a predetermined means (S201).
  • One music piece ID is assigned to each stored music piece data.
  • the impression value calculation unit 106 calculates an impression value for each piece of stored music data (S202).
  • the music signal of the stored music data is analyzed to calculate the feature amount of the music.
  • the calculated feature amount is weighted and added to calculate an impression value.
  • the impression value calculation unit 106 stores the calculated impression value in the impression value database 104 in association with the music piece ID of the music data of the calculation source (S203).
  • the music database 102 and the impression value database 104 necessary to search for music data using the impression values are generated.
  • a music information database 103 is also created as needed when music data is stored.
  • the conversion unit 110 acquires an impression value of the music to be searched (S 204).
  • an impression value of the music to be searched S 204.
  • the search result storage unit When performing a search to narrow down the results of the search performed immediately before, the search result storage unit
  • the song ID of the song data corresponding to the search performed immediately before is acquired from 101. Furthermore, the impression value corresponding to the acquired music ID is read out from the impression value database 104.
  • the determination as to whether the search target is the music data group narrowed down by the previous search or the power of all music is performed based on a predetermined operation of the user. This determination may be performed, for example, after the reception unit 107 receives a predetermined operation of the user, and the determination result may be notified to the range change unit 109 via the conversion unit 110.
  • Range changing unit 109 reads the impression word conversion table from conversion table storage unit 108.
  • a change is added to the read impression word conversion table according to the distribution of the impression values of the music to be searched (S 205).
  • the changing method is as described above, and the equation set in the impression word conversion table is changed according to the distribution of the impression value.
  • the receiving unit 107 receives an impression word specified by the user (S206).
  • the conversion unit 110 uses the impression word conversion table changed by the range change unit 109, and converts the impression word acquired by the reception unit 107 into a mathematical expression indicating an impression word range (S207).
  • the search unit 111 performs a search based on the music ID of the music to be searched and a mathematical expression indicating an impression word range.
  • the search is executed (S208). Specifically, the search unit 111 receives the music ID of the search target music from the range change unit 109 via the conversion unit 110. If the search target is a music group corresponding to the search performed immediately before, the device ID may be acquired from the result search result storage unit 101.
  • the search unit 111 receives from the conversion unit 110 a mathematical expression indicating an impression word range.
  • the search unit 111 reads out an impression value corresponding to the received music ID from the impression value database 104. Intermediate strength of read out impression value Search for something within the impression word range shown in the formula.
  • the search unit 111 stores, in the search result storage unit 101, the music piece corresponding to the search, that is, the information such as the music piece ID corresponding to the impression value present in the impression word range (S209).
  • the search unit 111 notifies the output unit 120 of the music ID of the music that corresponds to the search.
  • the output unit 120 reads out each of the music information and the music data corresponding to the notified music ID from the music information database 103 and the music database 102 as needed, and outputs it to the user (S210).
  • FIG. 7 is a conceptual view of an impression space showing a difference between a search result by the conventional music search system and a search result by the music search device 100 of the present embodiment.
  • the left drawing of FIG. 7 shows the search result by the conventional music search system
  • the right drawing of FIG. 7 shows the search result by the music search device 100 of the present embodiment.
  • the music to be searched is the music of a specific singer, and the distribution of the impression values is biased to the central part of the impression space. Therefore, when searching with the conventional music search system, as shown in the left figure of FIG. 7, a large number of musics correspond to "pop", and there is no music corresponding to "heal".
  • the music search device 100 changes the impression word range according to the distribution of the impression values. Specifically, the impression word range is changed by performing a linear mapping on the impression word range from the impression value space to the distribution range of the impression value. That Therefore, as shown on the right side of Fig. 7, there are appropriate numbers of music for "pop" and "healing".
  • the music search device 100 of the embodiment of the present invention even if there is a bias in the distribution of the impression values of the music piece to be searched, the distribution of those impression values. With the relative relationship between the distribution range of the impression value and the search word range changed according to, the search by the impression word becomes possible. As a result, an appropriate number of search results can be output even when any impression word is selected as a search word. As a result, it becomes possible to narrow down the search object appropriately.
  • the range changing unit 109 uses a map so that the impression word range in the impression space matches the distribution range of the impression values of the search target music group. To change it.
  • the inverse mapping may be used to change the impression values so that the distribution range of the impression values is scattered throughout the impression space. That is, if the relative positional relationship between the distribution range of impression values and the impression word range is changed according to the distribution of impression values, the distribution range of impression values may be changed.
  • FIG. 8 is a functional block diagram showing a functional configuration of the music search device 100 including the impression value changing unit 112. As shown in FIG.
  • the music search device 100 shown in FIG. 8 has the same effects as the music search device 100 according to the above-described embodiment. That is, even if there is a bias in the distribution of impression values for the search target music, the search target can be appropriately narrowed down.
  • the impression value changing unit 112 is provided to change the distribution range of the impression values in the impression space.
  • the impression value changing unit 112 is another example of the changing unit in the content data search device of the present invention.
  • the impression value of each music piece to be searched is changed by inverse mapping transformation which is inverse transformation of the mapping transformation shown in (Expression 2) and (Expression 3) described above.
  • (Eq. 5) and (Eq. 6) are inverse mapping transformation functions of (Eq. 2) and (Eq. 3). That is, (Eq. 5) and (Eq. 6) are functions representing a linear mapping of the impression value to the distribution range force impression value space.
  • FIG. 9 is a diagram showing an example of a process flow of the music search device 100 including the impression value changing unit 112.
  • conversion unit 110 reads the impression word conversion table from conversion table storage unit 108.
  • the receiving unit 107 receives an impression word designated by the user (S302).
  • the conversion unit 110 uses the impression word conversion table read out from the conversion table storage unit 108, and converts the impression word acquired by the reception unit 107 into a mathematical expression indicating an impression word range (S303).
  • the impression value changing unit 112 acquires an impression value of the search target music (S304). Specifically, when performing a search for narrowing down the search result performed immediately before, from the search result storage unit 101, the music ID of the music corresponding to the search performed immediately before is acquired. Furthermore, the impression value corresponding to the acquired music ID is read out from the impression value database 104.
  • the impression value changing unit 112 acquires all the impression values and music IDs stored in the impression value database 104. Do. It is determined based on a predetermined operation of the user whether the search target is the power of the music narrowed down by the immediately preceding search or whether the music is all music. This determination may be performed, for example, when the receiving unit 107 receives a predetermined operation of the user, and notifies the impression value changing unit 112 of the determination result via the converting unit 110 and the searching unit 111.
  • the impression value changing unit 112 changes the impression value of each music piece to be searched according to the above (Equation 5) and (Equation 6).
  • the search unit 111 executes the search based on the music ID of the music to be searched and the mathematical expression indicating the impression word range (S306). Specifically, the search unit 111 receives, from the impression value changing unit 112, the song ID of the song to be searched and the impression value after the change. Also, it receives from the conversion unit 110 an expression that indicates an impression word range. The search unit 111 searches the received impression values for those within the impression word range indicated in the formula.
  • the search unit 111 causes the search result storage unit 101 to store information corresponding to the search, ie, the music ID corresponding to the impression value present in the impression word range (S 307).
  • the search unit 111 notifies the output unit 120 of the music ID of the music that corresponds to the search.
  • the output unit 120 reads out each of the music information and the music data corresponding to the notified music ID from the music information database 103 and the music database 102 as needed, and outputs it to the user (S308).
  • the music search device 100 including the impression value changing unit 112 can narrow down the search target appropriately.
  • FIG. 10 shows the difference between the search result when the distribution range of the impression values is not changed and the search result when the distribution range of the impression values is changed by the flow of processing shown in FIG. It is a conceptual diagram.
  • the music to be searched is a music group narrowed down by a search with a specific singer name as in the case described with reference to FIG.
  • the distribution of impression values is biased to the central portion of the impression space. Therefore, when a search is made without changing the distribution range of the impression value, a large amount of music corresponds to "pop" and no search corresponds to "heal". This is a process performed by a conventional music search system.
  • the music search device 100 including the impression value changing unit 112 changes the distribution range of the impression values. Specifically, the impression value is changed by transformation of the distribution range force of the impression value by linear mapping to the impression value space. Therefore, as shown on the right side of Fig. 10, the appropriate number of songs correspond to both "pop" and "healing".
  • the music search device 100 includes the impression value changing unit 112 for changing the distribution range of the impression values as shown in FIG. 8 instead of the range changing unit 109 for changing the impression word range. Good.
  • the music search device 100 may include the impression value changing unit 112 together with the range changing unit 109 instead of the range changing unit 109 in addition to the impression value changing unit 112.
  • the range changing unit 109 may change the impression word range.
  • the impression value changing unit 112 may change the impression value.
  • both the impression value and the search term range may be changed.
  • a reference two-dimensional space having a predetermined size is determined.
  • the impression value is changed by linear mapping so that its distribution range has the same size as the reference two-dimensional space.
  • the search word range is changed by linear mapping from the impression word space to the reference two-dimensional space.
  • the search target can be appropriately narrowed down.
  • either one of the impression value and the search term range may be changed. You may change both. By doing this, for example, according to the number of tunes to be searched and the number of impression words set in the impression word conversion table, the change processing with less processing load is performed. You can choose the law.
  • a plurality of types of impression word conversion tables may be prepared in advance, and the range changing unit 109 may select an impression word conversion table to be used for the search according to the characteristics of the music piece to be searched. Yes.
  • an impression word conversion table for each artist is stored in the conversion table storage unit 108 according to the distribution of impression values of music pieces for each artist.
  • the range change unit 109 selects an impression word conversion table corresponding to the artist if the music group to be searched is a music group narrowed down by a search by a specific artist name. Thereafter, a search is performed using the selected impression word conversion table.
  • the selected impression word conversion table a plurality of post-impression ranges are set in advance according to the distribution of the impression values of the search target music. Therefore, preparing the impression word conversion table for each artist has the same effect as changing the impression word range according to the distribution of the impression value of the search target music by the range changing unit 109. That is, even if the distribution of impression values of the music to be searched is biased, the search can be narrowed down appropriately.
  • the artist name stored in the music information database 103 as the range changing unit 109 and the artist name stored in the impression value database 104 Get the impression value of the song and create it in advance. /.
  • an impression word conversion table for different genres of music may be prepared for each artist. It is considered that the impression value of music often depends on the genre of music. Therefore, in the same way as the above-mentioned artist-specific impression word conversion table, a genre-specific impression word conversion table is prepared.
  • Range change section 109 selects an impression word conversion table corresponding to the genre when the music group to be searched is a music group narrowed down by a search with a specific genre name. Thereafter, a search is performed using the selected impression word conversion table.
  • the form in which the impression word conversion table for each genre is prepared has the same distribution as the form in which the impression word conversion table for each artist is prepared, the distribution of the impression values of the search target music Even if there is a bias, it is possible to narrow down the search object appropriately.
  • the genre of impression word conversion table is, for example, the genre name stored in the music information database 103 by the range changing unit 109 and the music belonging to the genre stored in the impression value database 104. Get an impression value and create it with confidence.
  • the range changing unit In 109 by preparing the impression word conversion table for each artist or genre in advance, if the search target is a music group narrowed down with a specific artist name or genre name, the range changing unit In 109, there is no need to perform calculations to change the impression word range. That is, it is possible to reduce the burden of processing at the time of search.
  • Each of the artist name and the genre name described above is an example of an attribute for dividing a plurality of music pieces into groups, and may be a label name to which the music pieces belong, a year of release of the music pieces, or the like.
  • an impression word conversion table for each label is prepared.
  • search target is a group of songs narrowed down with a specific label name, then search using the impression word conversion table corresponding to that label name.
  • range changing section 109 changes the size and position of the impression word range using (Equation 2) and (Equation 3) described above. It is possible to change only the size while pushing.
  • the impression word range set in the impression word conversion table in the present embodiment is a circle, only the radius may be changed without changing the center of the circle.
  • FIG. 11 is a conceptual diagram showing an example where only the radius of the impression word range is changed in the impression space.
  • the number of impression values included in the impression word range of the impression word “healing system” is set in the impression word conversion table, the number of impression values included is greater than or equal to a predetermined number. Expand the range of impression words until For example, the impression word range is expanded to include three or more. In this case, at least three impression values close to the central force of the impression word range will be selected. Therefore, even in a mode in which the impression word range is expanded as needed without changing the center position of the impression word range, the search target can be appropriately narrowed down.
  • the range changing unit 109 continues the process until the impression value is included in a predetermined number or less.
  • the range may be reduced. For example, if the impression word category includes an impression value of 6 or more, the impression word range may be reduced to 5 or less! / ⁇ .
  • the search target can be appropriately narrowed down.
  • the image word range may be expanded by reducing or expanding the image word simply by expanding and contracting the radius, and may be changed into an ellipse according to the distribution of the image values that are not.
  • the circle that is the impression word range is changed to an ellipse in which only the radius in the Y-axis direction is extended. It may be scaled up or down with.
  • the ratio of the length of the impression word range in the Y-axis direction to the length in the X-axis direction was modified to be equal to the ratio of the length in the Y-axis direction of the distribution range of impression values to the length in the X-axis direction. You can zoom in or zoom out on.
  • the function of expanding or reducing the impression word range may be added to the form of changing the impression word range using (Equation 2) and (Equation 3) described above. Also, it may be added to the form of changing the impression value using (Equation 5) and (Equation 6) above! / ,.
  • the impression space is two-dimensional.
  • the impression space may be one dimension or three dimensions or more.
  • the range changing unit 109 or the impression value changing unit 112 may change the impression word range or the impression value according to the distribution of the impression value in the three-dimensional impression space, using the mapping function or the inverse mapping function.
  • the search term is an impression term that is a term that intuitively expresses music, such as “Norinori type”, “Pop type”, “Healing type” and the like.
  • fuzzy words other types of vague words (hereinafter referred to as “fuzzy words”) may also be used.
  • the attribute of the music indicating the announced age, age, or period of the music, etc. " Recent ,, and, "Old ,, and so on!
  • Ambiguous words such as “recent” and “old” can also be replaced with attribute values indicating the published age, age, or period of the music, but from these vague words
  • the absolute age, etc. to be associated differ depending on the distribution range on the time axis of the music piece to be searched.
  • the attribute value indicating the announced age, age, period, etc. of the music is hereinafter referred to as the “year value”.
  • the music search device 100 can be applied to a mode in which music data is searched using such fuzzy words as search words, and narrowing down the search target appropriately Can.
  • the attribute value associated with the music is one kind of age value indicating the released age of the music
  • the space in which the age values are distributed is a one-dimensional space.
  • this one-dimensional space is referred to as the “period space”.
  • the search term range is the range of age values corresponding to v, words in the age space.
  • FIG. 12 shows the difference between the search result with a certain search term range and the search result with a search term range that has been changed according to the distribution of age values, when music data is searched with fuzzy words related to the age.
  • the left side of FIG. 12 shows the search result by a certain search term range
  • the right side of FIG. 7 shows the search result by the search term range changed according to the distribution of age values.
  • the left figure in Fig. 12 is the search result when the conventional music search system is applied to this fuzzy word search.
  • the search target is the music groups of singer A and singer B respectively.
  • the age space has a predetermined size.
  • the predetermined size is, for example, a period such as 1990, 2005 until 2005.
  • the search word range of “recent” is constant. That is, in the prior art, search The word range is fixed at a fixed period based on, for example, the present or a predetermined year and month. Therefore, for singer A, there is no corresponding song for force singer B, which has three songs.
  • the music search device 100 can change the search term range according to the distribution of the age values in the age space. It is possible to output an appropriate number of search results for both the singer and singer B as well.
  • the conversion table storage unit 108 stores a search term conversion table in which a term such as an age is shown in association with a range of term values. For example, the information showing up to 6 months ago is displayed in association with "recent".
  • an age value database is installed in the music data storage unit 105, and the age value of each music is stored in association with the music ID.
  • Range changing section 109 changes the search term range in accordance with the distribution of age values in the age space of the music piece to be searched.
  • the search term range of "recent" is changed by linear mapping from the age space to the distribution range of the age values.
  • the music search device 100 can appropriately narrow down the search targets even when there is a bias in the distribution of age values of the music pieces to be searched.
  • the distribution range of age values and the search word range are relative to each other by changing the age value which is not the same as changing the search word range in the same manner as searching music data by impression words. Even if the relationship is changed, the above-described appropriate narrowing effect is exhibited.
  • the power described for the music search device 100 for searching music data may be content data other than music data to be a search target.
  • it may be moving image data or still image data! / ⁇ .
  • at least the impression values of the moving image may be prepared as a database.
  • the ability to change the distribution range of attribute values such as impression word range or impression value, which is the feature of the present invention, is not lost even if the search target is any type of data.
  • a search may be performed using a word or phrase indicating an attribute such as an impression word. It is possible to narrow down the search target appropriately by changing the impression word range mentioned above.
  • an impression word used as a search word may be associated with a mathematical expression or the like indicating an impression word range which is the same as that shown in FIG.
  • the impression words shown in FIG. 4 may not be used as they are for the user to select.
  • the impression word range corresponding to the impression word selected by the user may be specified by the impression word conversion table.
  • the impression word conversion table may be one that associates an impression word with an impression value represented by one point.
  • the impression word range which is the range of the impression value corresponding to the impression word may not be the mathematical expression itself as shown in FIG.
  • information indicating the center and radius of a circle may be used to specify a search term range.
  • the information indicating the changed impression word range passed to the search unit 111 by the conversion unit 110 may be information indicating the focal point and the length ratio of an ellipse instead of the equation itself. In this way, if the information that indicates the search term range is information that can uniquely identify the scope within the impression space even if it is not an equation, it may be.
  • linear mapping conversion is performed.
  • non-linear transformations may be used instead of linear transformations.
  • multiple linear and non-linear functions may be used according to the distribution of impression values.
  • At least the impression word range may be reduced and moved so as to be included in the distribution range of the impression value.
  • the relative relationship between the distribution range of the impression value and the impression word range can be changed in the direction in which the narrowing effect of the search is exhibited.
  • an impression word designated by the user may be received (S207), and then an impression value may be acquired (S204).
  • the range changing unit 109 may read and change only the mathematical expression indicating the search term range corresponding to the search term received by the receiving unit 107.
  • the order of processing relating to the search execution in the music search device 100 and the process contents necessary for the search execution are not limited to the order and the process contents in the above-described embodiment.
  • the relative relationship between the distribution range of the impression value and the impression word range is changed according to the distribution of the impression values of the music piece to be searched, and the user is designated by the user in the relative relationship after the change. It is sufficient if it is possible to search music data by the search word. Therefore, the order of processing relating to search execution may be determined by the user's needs or the actual value of processing speed.
  • the music ID is used as an identifier for identifying the music, it is possible to identify the music, and if it is information that can associate music data, music information and an impression value, for example, It may be a song number or the like.
  • the music data storage unit 105 provided in the music search device 100 has a music database 10.
  • the music data storage unit 105 stores a plurality of music data, music information, and an impression value in association with each other, the music data storage unit 105 may be configured as a single database.
  • the impression value database 104 stores the impression value of each music calculated by the impression value calculation unit 106, and then the search is executed. I did.
  • the impression values stored in the impression value database 104 those which have been calculated may be used.
  • the impression value database 104 may be generated by causing the music search device 100 to read an impression value that is also required for storage medium power where impression values of a plurality of songs are stored! /.
  • the music data and the music information may be stored in an external database.
  • It has a database in which at least the impression values and the music IDs are associated and stored Then, it is possible to identify the music corresponding to the search.
  • the search result storage unit 101 is assumed to store information on the search results in the music search device 100.
  • information related to the results searched by another search device or the like, for example, the music ID of the music corresponding to the search may be stored. That is, the music search device 100 can further narrow down the search targets narrowed down by another search device or the like.
  • output unit 120 outputs the music ID, the music data, and the like corresponding to the search to the user.
  • the display and playback of the music ID, music data, etc. to the user may be performed by another device.
  • the music search device 100 may be incorporated in a network distribution system that distributes music data via a network such as the Internet and used for searching music data. As described above, the music search device 100 can appropriately narrow down the search targets. Therefore, it is suitable for a network distribution system in which it is necessary to efficiently search out music data desired by the user from a large amount of music data.
  • a playlist may be generated in which the song ID of the song data corresponding to the search and the titles of the songs included in the song information are listed.
  • the method and method of outputting search results to the user may be designed according to the user's request, manufacturing cost, and the like.
  • the content data search device of the present invention can change the relative relationship between the distribution range of content attribute values and the search word range according to the distribution of content attribute values. it can. Therefore, even if the distribution of attribute values of search target content is biased, it is possible to appropriately narrow the search target.
  • the content data search device of the present invention is an audio device and a car navigation system with an audio function.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Computational Linguistics (AREA)
  • Library & Information Science (AREA)
  • Multimedia (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

Système de recherche de données de contenu pouvant cribler de façon adéquate des objets de recherche même si la distribution des valeurs d’attribut des contenus objet de la recherche est irrégulière. Le système de recherche de données comprend un base de données de valeur d’impression contenant une valeur d’attribut, une table de conversion où un mot recherché est mis en relation avec un plage de mots recherchés qui est la plage de valeurs d’attribut correspondant au mot recherché, une section de modification de plage pour modifier la relation relative entre la plage de distribution des valeurs d’attribut et la plage des mots recherchés en modifiant la lecture de la plage des mots recherchés depuis la table de conversion, une section de réception pour recevoir l’entrée d’un mot recherché et une section de recherche pour rechercher les données de contenu correspondant au mot recherché reçu par la section de réception en cherchant à partir de la valeur d’attribut incluse dans la plage de mots recherchés correspondant au mot recherché par rapport à la relation relative entre la plage de distribution des valeurs d’attribut modifiée par la section de modification de plage et la plage de mots recherchés.
PCT/JP2005/018353 2004-11-04 2005-10-04 Système de recherche de données de contenu WO2006048988A1 (fr)

Priority Applications (4)

Application Number Priority Date Filing Date Title
US11/665,663 US7743049B2 (en) 2004-11-04 2005-10-04 Content data retrieval apparatus
JP2006542314A JP4659755B2 (ja) 2004-11-04 2005-10-04 コンテンツデータ検索装置
EP05790249.6A EP1832987B8 (fr) 2004-11-04 2005-10-04 Système de recherche de données
CN2005800368296A CN101048776B (zh) 2004-11-04 2005-10-04 内容数据检索装置

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2004320198 2004-11-04
JP2004-320198 2004-11-04

Publications (1)

Publication Number Publication Date
WO2006048988A1 true WO2006048988A1 (fr) 2006-05-11

Family

ID=36319003

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2005/018353 WO2006048988A1 (fr) 2004-11-04 2005-10-04 Système de recherche de données de contenu

Country Status (5)

Country Link
US (1) US7743049B2 (fr)
EP (1) EP1832987B8 (fr)
JP (1) JP4659755B2 (fr)
CN (1) CN101048776B (fr)
WO (1) WO2006048988A1 (fr)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2009110605A1 (fr) * 2008-03-07 2009-09-11 日本ビクター株式会社 Dispositif serveur, dispositif terminal, dispositif de reproduction
JP2009216812A (ja) * 2008-03-07 2009-09-24 Victor Co Of Japan Ltd サーバ装置、端末装置、再生装置、分類構造データ送信方法、コンテンツ再生方法、コンピュータプログラム
JP2009223048A (ja) * 2008-03-17 2009-10-01 Victor Co Of Japan Ltd サーバ装置、端末装置、再生装置、プレイリスト構造データ送信方法、コンテンツ再生方法、コンピュータプログラム

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2007172702A (ja) * 2005-12-20 2007-07-05 Sony Corp コンテンツ選択方法及びコンテンツ選択装置
JP4683290B2 (ja) * 2006-01-23 2011-05-18 ソニー株式会社 表示装置、表示方法、表示プログラム
JP5135931B2 (ja) * 2007-07-17 2013-02-06 ヤマハ株式会社 楽曲加工装置およびプログラム
US8825646B1 (en) * 2008-08-08 2014-09-02 Google Inc. Scalable system for determining short paths within web link network
JP5458880B2 (ja) * 2009-03-02 2014-04-02 富士通株式会社 文書検査装置、コンピュータ読み取り可能な記録媒体および文書検査方法
GB201708771D0 (en) * 2017-06-01 2017-07-19 Upshoot Ltd Improved search engine

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH11161659A (ja) * 1997-11-27 1999-06-18 Nippon Telegr & Teleph Corp <Ntt> 検索式の動的変更方法及び装置及び検索式の動的変更プログラムを格納した記憶媒体
JP2001022778A (ja) * 1999-07-09 2001-01-26 Fuji Xerox Co Ltd 検索装置及び検索方法
JP2001134573A (ja) * 1999-11-05 2001-05-18 Nippon Telegr & Teleph Corp <Ntt> 類似データ検索方法，類似データ検索装置，および類似データ検索用プログラム記録媒体
JP2003016078A (ja) * 2001-06-29 2003-01-17 Hitachi Ltd データの分類、検索および表示方法

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004086189A (ja) 2002-07-02 2004-03-18 Matsushita Electric Ind Co Ltd 楽曲検索システム
EP1378912A3 (fr) * 2002-07-02 2005-10-05 Matsushita Electric Industrial Co., Ltd. Système de recherche de musique

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH11161659A (ja) * 1997-11-27 1999-06-18 Nippon Telegr & Teleph Corp <Ntt> 検索式の動的変更方法及び装置及び検索式の動的変更プログラムを格納した記憶媒体
JP2001022778A (ja) * 1999-07-09 2001-01-26 Fuji Xerox Co Ltd 検索装置及び検索方法
JP2001134573A (ja) * 1999-11-05 2001-05-18 Nippon Telegr & Teleph Corp <Ntt> 類似データ検索方法，類似データ検索装置，および類似データ検索用プログラム記録媒体
JP2003016078A (ja) * 2001-06-29 2003-01-17 Hitachi Ltd データの分類、検索および表示方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP1832987A4 *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2009110605A1 (fr) * 2008-03-07 2009-09-11 日本ビクター株式会社 Dispositif serveur, dispositif terminal, dispositif de reproduction
JP2009216812A (ja) * 2008-03-07 2009-09-24 Victor Co Of Japan Ltd サーバ装置、端末装置、再生装置、分類構造データ送信方法、コンテンツ再生方法、コンピュータプログラム
CN101990766B (zh) * 2008-03-07 2013-04-03 日本胜利株式会社 服务器装置、终端装置、再生装置
US8554880B2 (en) 2008-03-07 2013-10-08 JVC Kenwood Corporation System for classification, location and transmission of music playlists
JP2009223048A (ja) * 2008-03-17 2009-10-01 Victor Co Of Japan Ltd サーバ装置、端末装置、再生装置、プレイリスト構造データ送信方法、コンテンツ再生方法、コンピュータプログラム

Also Published As

Publication number Publication date
EP1832987A4 (fr) 2009-12-09
EP1832987A1 (fr) 2007-09-12
CN101048776A (zh) 2007-10-03
US7743049B2 (en) 2010-06-22
JP4659755B2 (ja) 2011-03-30
CN101048776B (zh) 2010-08-18
EP1832987B8 (fr) 2016-10-12
US20080222128A1 (en) 2008-09-11
JPWO2006048988A1 (ja) 2008-05-22
EP1832987B1 (fr) 2016-04-13

Similar Documents

Publication Publication Date Title
JP4659755B2 (ja) コンテンツデータ検索装置
US8166044B2 (en) Method and apparatus for selecting items from a number of items
US7840559B2 (en) User terminal and content searching and presentation method
US7805389B2 (en) Information processing apparatus and method, program and recording medium
US8073854B2 (en) Determining the similarity of music using cultural and acoustic information
US7842876B2 (en) Multimedia object grouping, selection, and playback system
US20070106405A1 (en) Method and system to provide reference data for identification of digital content
US20100274782A1 (en) Generating metadata for association with a collection of content items
US20100217755A1 (en) Classifying a set of content items
JPH06110945A (ja) 音楽データベース作成装置及びその検索装置
KR100978689B1 (ko) 미디어 선택 방법 및 시스템
JP2000155759A (ja) 音楽情報検索装置，音楽情報蓄積装置，音楽情報検索方法，音楽情報蓄積方法およびそれらのプログラムを記録した記録媒体
JP2001306580A (ja) 音楽データベース検索装置
Gurjar et al. Comparative Analysis of Music Similarity Measures in Music Information Retrieval Systems.
JP5516720B2 (ja) カテゴリ生成プログラム、カテゴリ生成装置およびカテゴリ生成方法
JP2003316818A (ja) 情報検索方法及びその装置、コンピュータプログラム
JP2003178088A (ja) プレイリスト作成装置およびプレイリスト作成方法、情報再生装置、並びに、プログラム記録媒体
JP2010165160A (ja) 楽曲分類装置、楽曲検索システム及びコンピュータプログラム
JPH02136971A (ja) マルチメディア・データベースにおける検索方法
JP2007241941A (ja) コンテンツ検索装置及びコンテンツ検索プログラム
JP5521094B2 (ja) 楽曲分類装置および楽曲検索システム
JP2002007442A (ja) データ検索装置、データ検索方法及びデータ検索プログラムを記録した記録媒体
Hoashi et al. Comparison of User Ratings of Music in Copyright-free Databases and On-the-market CDs
JP2006293896A (ja) 楽曲検索装置
JP2006146586A (ja) 検索データベース作成装置、情報検索装置および情報検索システム

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BW BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE EG ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KM KP KR KZ LC LK LR LS LT LU LV LY MA MD MG MK MN MW MX MZ NA NG NI NO NZ OM PG PH PL PT RO RU SC SD SE SG SK SL SM SY TJ TM TN TR TT TZ UA UG US UZ VC VN YU ZA ZM ZW

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): GM KE LS MW MZ NA SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IS IT LT LU LV MC NL PL PT RO SE SI SK TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG

WWE Wipo information: entry into national phase

Ref document number: 2006542314

Country of ref document: JP

121 Ep: the epo has been informed by wipo that ep was designated in this application
WWE Wipo information: entry into national phase

Ref document number: 11665663

Country of ref document: US

WWE Wipo information: entry into national phase

Ref document number: 200580036829.6

Country of ref document: CN

WWE Wipo information: entry into national phase

Ref document number: 2005790249

Country of ref document: EP

NENP Non-entry into the national phase

Ref country code: DE

WWP Wipo information: published in national office

Ref document number: 2005790249

Country of ref document: EP